Masayoshi Son is one of AI's most steadfast believers. His SoftBank has poured nearly $65 billion into OpenAI, and in recent years, he has repeatedly emphasized the returns and possibilities AI brings.

But on Sunday, at the "Tech and Society" forum in Kyoto, he spoke unusually candidly about the risks.

"We no longer have the luxury of letting humanity fight among ourselves," he said on stage alongside Michael Kratsios, the White House's technology policy advisor. With AI capabilities growing exponentially, countries must build mutual trust to navigate this technology together. Superintelligence, if it falls into the wrong hands, would become "super dangerous."


Coming from Son, these words carry more weight than from most people. In recent months, safety incidents involving AI models have piled up, and public concern has intensified. A man who bet $65 billion on AI is now publicly flagging risks—that in itself is a signal.

Kratsios came to Kyoto to announce a U.S.-led cooperation plan involving 17 countries to accelerate scientific research with AI. Representatives from Japan, South Korea, the UK, Germany, Singapore, the UAE, and other nations joined him on stage.

Cooperation aside, there was another layer to Son's remarks: as models grow more powerful, whether safety mechanisms can keep up is a real question facing every country. After delivering his warning, he turned around and continued running his AI empire. That's the part worth pondering most.

(Compiled from Global Market Broadcast)